In this contribution, the magnetic characterisation of steel strips is studied using synthetic
data of field-gradient transients, which have been produced via the finite integration
technique (FIT). The material law is described and parametrized using the Jiles-Atherton (JA)
model. The sensitivity of relevant magnetic indicators with respect to the material parameters
is then analyzed using two global methods: Sobol indices and $\delta$-sensitivity
indices. In order to accelerate the evaluation of these quantities, a fast metamodel is built
using machine learning techniques from a simulated dataset. The solution of the inverse
problem based on a tailored learning framework is tested for the different proposed
identifiers, and their suitability for the magnetic characterisation of the material in
question is finally discussed.